Work

Example ADR Review Prompt and corresponding LLM-based ADR Reviews

Year: 2026

Type: dataset

Abstract: This dataset contains a review prompt for an exemplary architectural decision record (ADR) and the corresponding reviews/evaluations done by different LLMs. This dataset accompanies the short paper "AI Assistance for Architectural Decision Making: Domain-Driven Context and Prompt Engineering" submitted to the IDEA-ARCH workshop @ ECSA 2026. The *.txt file contains the specific review prompt including the specific ADR to review. The *.yaml files contain the structured reviews based on this prompt made by different Large Language Models. Their filenames indicate the model name, version, and if they were run locally, including the number of parameters.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.20207403

Authors: Zimmermann, Olaf, Seiger, Ronny

Institutions: University of St.Gallen, RWTH Aachen University

Language: English

Topic: Software Engineering Techniques and Practices

Subfield: Information Systems

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funder: Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung

Award: Towards Sustainable Software Architectures for Cyber-physical Systems of Systems (TaSSAreCt)

OpenAlex ID: W7161277413

Cites:

Cited by:

Related to: